Don't ever consent to anything if you are the target of a criminal investigation or anything (like a traffic stop) that might turn into a criminal investigation. What you think they're going to do with your consent and what they intend to do may be two different things. You may be surprised at what the courts *allow* them to do with information you've given them permission to access it.

Be scrupulously cooperative with anything the cop is allowed to demand that you do. Don't argue, lecture or harangue, do everything you can to make an encounter with the police, smooth, courteous and above all, brief. But politely and firmly draw the line if they ask you for information about yourself other than your identity (which in some jurisdictions you are required to provide). Do not resist, because that can get you tased. Cooperate, but make it clear you don't consent, then sort things out later with your lawyer.

Be on your guard especially with a polite cop. Many of them are polite because they were brought up that way, but the smart ones know that courtesy is a powerful way to gain cooperation. They get you into a rhythm by politely asking for a series of reasonable things, then slipping in a request for something they can't compel, like you opening your trunk, granting them access to your cell phone, or letting them search your house.

It would be great if all cops abided by Peelian principles, but the "War on Drugs" has undermined the relationship between the police and the public. Fully ONE PERCENT of all Americans are behind bars, most of them on drug charges. If you, any of your family, or any of your friends have so much as smoked a joint, you can't afford to give the police any personal information at all.

You don't get it: inciting impotent rage is a *feature*. When you feel rage, everything becomes clear. That's *so* much more satisfying than gnawing, existential doubt.

Because this method returns a ranking that seems more intuitively "right"?

It's worth asking why that should be. Think about a rating scale of one to four stars. What does *averaging* those ratings mean? Yes, I know the *formula* for computing an average, but being able to *compute* an average isn't the same thing being able to *interpret* that average. Why? Because a two star rated item isn't really "twice as good" as one star, a four star rating isn't really "four times as good" as a one star or "twice as good" as a two star. In other words we're not measuring something like height, or number of widgets sold. We're *characterizing* something as shit, meh, OK, or awesome.

The method proposed is more mathematically defensible. It converts each 1 to 4 score into a coin flip (Bernoulli process for you probability geeks): positive or negative. This actually captures what the rating system does better than treating the ratings as meaningful numbers.

This doesn't automatically mean you'll get more intuitively pleasing results than using weighted averages; YMMV. But I think this system is bound to give more *consistently* intuitive results across applications and datasets. Weighting the average is an easy, but extremely arbitrary operation. As a compromise you might come up with a formula to tweak the score *toward the mean* depending on how few reviews there were; by adjusting the parameters you use you could probably get that formula to yield an intuitively satisfying ranking. The advantage of using a confidence interval as suggested is that approach is highly likely to yield a reasonable result out of the box, across a wide variety of datasets.

And after all, even if it looks like a "hot mess" to people who haven't taken advanced statistics, how hard is it to copy and paste a bit of code?

It's not an algorithm, except in the trivial sense. It's a formula for calculating an adjusted rating value that discounts extreme ratings for items with small numbers of reviewers.

This actually matches what you do intuitively when you see an item with a single rating of 5.0 at the top of a list, just above another item with an average rating of 4.9 from a thousand users. You mentally deduct a bit from the "top rated" item because you know it's probably too high. Likewise a 1.0 rating from a single user is probably too low, so you mentally add a bit to that.

The question is, how much to deduct or add from the score?

The approach suggested is to ask a slightly different question. Instead of "what is the average rating of the product", you ask "what percentage of positive ratings can I be 95% certain the product would score above have if *everyone* rated it?" It turns out there's a number of mathematical formulas that are supposed to tell you precisely that.

There's still a lot of arbitrariness in this approach. Why 95%? I'm reasonably sure that results would be just as intuitively reasonable if we chose 80% instead. But if 95% seems to generate intuitively reasonable results there's no particular reason to monkey with that parameter.

BUT, I think, the level of arbitrariness involved probably means we could choose a simpler approximation than the Wilson interval if we could dream one up. The more familiar Wald interval taught in basic statistics courses is somewhat simpler, but not so much that it's worth worrying about, at least not if you're doing the calculation on a database server which typically has a few CPU cycles to spare.

If I were to attempt something like this on a massive scale in an environment where CPU cycles were precious, I'd probably devise some kind of simple algebraic scaling formula that tweaked scores toward the mean, depending on the number of ratings. The results wouldn't be quite as good as the Wald or Wilson intervals, but maybe not so much less good that anyone would notice.

I suspect the restriction is impossible to enforce, because it's almost certainly the case that the facial recognition isn't performed on the device itself. So it's a bit like saying you can't use the things for pornography; you'd have to know somehow that the user intends to pleasure himself later by looking at pictures of ladies' shoes.

It's a bit too late on that score anway. Having boots on the ground is an anachronism, even if they've got high tech wearables. In 2000 Scotland Yard was able to foil the Millennium Dome diamond heist by tracking all the gang's preparations using a network of public and private security cameras. There was almost no in-person surveillance. The first time the gang was physically near the police was when they were surrounded by an armed response team inside the exhibit hall.

Here in the US, the NYPD has acquired similar, possibly even more advanced capabilities. They can for example find all the six foot-ish blond men wearing blue sweaters in a one block radius of Penn Station then run their images through facial recognition software, then follow their suspect almost anywhere in Manhattan with no chance of being detected.

In comparison to camera networks, having a cop wearing Google glasses isn't that big a deal. A lot of our legal privacy protections are based on the assumption it's too labor intensive to follow people around for frivolous purposes. An army of cops with cameras is expensive to maintain; a network of surveillance cameras is not.

You really think NASA wasn't bureaucratic back in its glory days?

There were two big differences between then and now: then NASA had big-time funding and a clear mission that was stable over the course of a decade.

Bureaucracy is the optimal design pattern for an organization that has to accomplish a fixed set of objectives within a predictable set of constraints and with a stable set of resources. When bureaucracy works well, it becomes practically invisible. For example one of the greatest innovations Scotland Yard introduced to crime fighting is record keeping. That enables them to correlate crimes with evidence from past crimes in their records, including fingerprints. Before Scotland Yard introduced filing to police work, cops didn't fill out reports, and if a criminal got away with a crime the detectival slate was wiped clean for his next crime. But we usually don't think of a well-functioning police force as a record-keeping organization. When it works, you don't notice it.

Bureaucracies under stress are very noticeable. That's why the military is notoriously bureaucratic, because the bureaucracy's need to make it's work predictable and repeatable conflicts with the fluid nature of war. So "bureaucracy" in our minds is associated with malfunctioning or even corrupt organizations.

Bureaucracies are predictable. If you underfund a bureaucracy, it will inevitably put more effort into preserving itself and less into what it's supposed to do. If you ask it to accomplish something the people working in it perceive as impossible, false or contrived information will inevitably clog its record keeping system. If you demand schoolteachers raise score levels, score levels will rise whether or not children are getting better educated. If you demand cops increase the numbers of arrests, the numbers of arrests reported will rise whether or not there is an impact on crime.

So what happened to NASA's bureaucracy? Well, in the glory days, the environment was optimal for the functioning of bureaucracy. Even disasters like Apollo 1 weren't "unpredictable" from a planning standpoint. Things like that were bound to happen. The instant one did, the bureaucracy leapt into action, forming committees and tasks forces, like white blood cells rushing to the location of a wound. But once we'd landed on the Moon, things changed. Funding dropped, and the bureaucracy was asked to deal with conflicting priorities. Conflicting priorities aren't one of the things bureaucracies can handle.

You can't ask a bureaucracy to fix itself. Fixes have to come from a higher level. A company with a malfunctioning management structure needs to be addressed by the board. A malfunctioning government agency has to be addressed at the political level. It's not bureaucracy per se that brought low, it's politicians who don't want to make tough decisions about NASA priorities, nor adequately fund all the agency's objectives.

The FCC certainly regulates the wi-fi spectrum, as you would know if you ever bothered to even glance at all the pieces of paper that come with anything operating in the 2.4 or 5 GHz band.

If those bands were actually unregulated, you'd be able to go on Amazon and buy a 1000 watt wi-fi booster.

Wasting power we're currently throwing away anyhow seems like an odd concern. You could even store the excess energy in a really inefficient storage medium and still come out ahead -- provided that excess power was cheap enough to produce in the first place and the storage mechanism was cheap enough. It's the *financial* return that will determine behavior; the physical efficiency is only a contributing factor.

The real limiting point for percentage generated by solar will be the result of a complicated mix of factors, including technologies, economies and diseconomies of scale, local supply and demand etc. Where diminishing returns will hit zero in thirty-five years is impossible to estimate today using back-of-the-envelope calculations. Thirty-five years is a long time for a technology just hitting its stride. Consider: 35 years ago was roughly when the 8088 CPU was launched.

Well, you can call it a "double standard", but that's how science works. Interpretations of data that contradict the established theory face higher burdens of proof than interpretations that support it.

It's frustrating to climate denialists and perpetual motion inventors that their ideas aren't given equal dignity with the scientific consensus, but that's because by in large they're ignorant of the effort that went into forging that consensus. It took fifty hard fought years for AGW to become the scientific consensus, and as a result it enjoys a privileged position: it gets to play the null hypothesis. To do otherwise isn't fair to the people who fought that fight for decades, and won. You can overturn the scientific consensus, but it's an uphill battle, as it should be.

Ebola is (according to the summary) completely contained in Nigeria and Senegal. This 2014 outbreak is all over West Africa, and according to TFA (I know, I know) the patient had just returned from Liberia, a West African country where the current outbreak has (obviously) not been contained.

Someone bringing this virus back is not so surprising. The big deal will be when we have our first case of endemic transmission -- when someone *catches* the virus here.

Oh, god. Mel Gibson's 1990 Hamlet was awful. It was the most asinine thing I've ever seen. Shakespeare for people who really *are* dummies. Reportedly it was director Franco Zeffirelli's attempt to make Shakespeare "less cerebral" and more accessible to the masses. What a choice to try that with! The whole point of Hamlet is that he's so damned smart the only person who can really stand in his way is him.

My point was that you've got to find an actor who can give a knowledgeable performance. Not some meat-head action star stunt cast miles out of his depth. I'd rather watch Arnold Schwarzenegger Hamlet.

I think the best film adaptation of Hamlet I've seen was Kenneth Branaugh's 1996 version, although it is long, long, long at 242 minutes (to Gibsons' 134 minutes). Olivier's 1948 Hamlet is generally highly regarded, but it's too sentimental for my taste. Haven't seen Derek Jacobi's 1980 BBC performance, but I've heard good things about it. I've seen snippets of the David Tennant Hamlet, and it looks promising, although it's hard to shake the impression that it's Dr. Who playing Hamlet.

Exactly. Security screwed up, and then they HAD to deal with it. It's not mere security theater to have a security checkpoint. Those checkpoints are demonstrably important.

Not many of us remember, but until 1973 there was no baggage screening, no metal detectors, and no id requirements for getting on a commercial flight. The number of skyjackings had climbed rapidly since the mid-50s so that in 1972 there were 11 skyjackings of commercial flights around the world, seven in the US.

After security checkpoints were introduced in the US, there wasn't another skyjacking in the US for three years. Then an occasional one now and then, as people found loopholes. There was one passenger airliner hijacking of a flight FROM the US in all the 1980s and none in the 1990s.

My conclusion is that the security measures put in place by 1990 were highly effective. 9/11 fit the pattern of the early dribs-and-drabs hijackings, the difference is Al Qaeda made an effort to do multiple simultaneous exploitations of the vulnerability they'd found. There hasn't been a hijacking of a US flight since then, but given that the last passenger hijacking BEFORE 9/11 was in 1987, it's likely that this long dry spell is mostly if not entirely due to banning blades from carry on luggage. That's not to say that EVERY other change since then is security theater. I think reinforcing cockpit doors and changing pilot training was a reasonable response. But a lot of the enhanced pat-downs, magic scanners, no-fly list shennanigans and such are no doubt bogus.

Your analysis depends on two assumptions. First, that at the daily peak the amount of solar produced exceeds the total demand for electricity. That's actually quite likely to happen in the long term in certain locations -- sunny, densely developed residential neighborhoods for example -- but not in others -- in a neighborhood that has a steel mill. Maybe in the short term in a few places if the adoption of rooftop solar accelerates even more.

One of the ways to alleviate this would be to improve the distribution grid so that the excess supply could be sold further away. But lets say the day comes that the peak solar production exceeds the total electricity demand. That brings us to the second assumption.

The second assumption is that electricity is charged at a flat rate all day long. Clearly if lots of excess solar is being produced at noontime, you could easily reduce the cost you charge to electricity consumers (or pay back to electricity). We already do peak vs. off peak rates for industrial users.

This combination of grid improvements and reduced peak rates will encourage people and businesses to concentrate their power usage around noon. Maybe you'll charge our electric car at a higher rate, or maybe even charge large industrial or household batteries. The losses hardly matter, since we were throwing away the sunshine anyway. Increased noon usage will offset the tendency for electricity rates to fall during peak generation periods.

Am I saying the utilities won't lose a little money in a few isolated spots in the short term? No. What I'm saying is that we're hardly facing some kind of insurmountable singularity. Certainly not any time soon, nor in the long term if we can bring ourselves to prepare for it.

I actually don't think that's true. I think what you're reacting to comes with the epic scale of the novel (SoI&F really is just one, long, continuous work) -- both in word count and the enormous cast of characters. It's a kind of literary clutter. If you boiled Game of Thrones down to the story of Ned Stark's rise and downfall, that would be quite a satisfying (although grim) story arc. The fact that the story goes on and on after that dissipates the emotional impact of that one story line.

At over 1.7 million words currently, Song of Ice and Fire is more than six times as long as typical English translations of the Illiad and Odyssey combined. Think about that. In the time it took you to read just the first volume of Song of Ice and Fire, you could have read BOTH the Illiad and the Odyssey. And as a bonus you'd have read BOTH the Illiad and the Odyssey.

As works go further and further north of 200,000 words, they almost inevitably lose the tight, clockwork structure you expect in a 2 hour stage play or 70,000 word novel. Stories stop feeling like they have a beginning, middle, and end and start to feel more episodic. That happens to some stories well before they hit the 200,000 word mark (American Gods, 183 KWords).

At 473 KWords, Lord of the Rings is one of the rare exceptions. From Rivendell onward it's a marvel of complex yet tightly interwoven structure. But it's a hot steaming mess of false starts up until Ford of Bruinen. Tom Bombadil anyone? I think that it could probably be edited down to 400,000 words without losing much artistically. That's still almost miraculously long for a story that feels like one story.

I have a theory about episodic megastories like Song of Ice and Fire, which is that they aren't catharsis you get from a tightly plotted play or novel. They're about transporting a reader to a world he finds interesting to visit again and again. If so that bodes ill for the the Game of Thrones TV series now that Emilia Clarke has sworn off nude scenes.

Some would say this doesn't deserved to be dignified with a response, but I disagree.

The best introduction to Shakespeares plays is to see them on stage, performed by actors who know how to perform Shakespeare. Because of the shift in language, there's special skill needed for presenting Shakespeare to modern audiences. You'll be amazed at how much you understand. Until you know the play's text you'll be missing a lot too, but in the performance you won't notice that.

I'd go so far as to say it's better to see a Shakespeare play performed first before attempting to read it. Then tackle the text with its footnotes on every line.